• The Fed's Balance Sheet May Remain Larger for Longer
  • The Currency of Currency
  • Getting a Grip on Gigs
  • Daydreaming recently at 30,000 feet, I remembered one of my father's favorite jokes. After a commercial jet takes off, an announcement comes over the intercom. "Ladies and gentlemen," it says, "this is the first passenger airplane directed by an autopilot program. It was developed by experts at IBM, and represents an exciting new application of computers. So with that, please sit back and enjoy your flight… enjoy your flight… enjoy your flight… enjoy your flight…"

    I'm not sure why my father found that so funny, as the apparent computer malfunction would likely have stricken the passengers with fear. Implicit in the telling, perhaps, was that an actual pilot was in the cockpit with the ability to override the program if it strayed from the intended course.

    A little over a year ago, the Federal Reserve published an outline of how it would go about reducing its balance sheet. The plan was activated last October. It called for a slow start and a gradual moderation to avoid an adverse economic reaction. Fed Chairman Jerome Powell recently reiterated the characterization his predecessor had offered: the effort is on "autopilot."

    But recent events have cast doubt on the Fed's ultimate destination. Operating challenges have arisen, prompting speculation that the Fed may conclude its unwinding program sooner than expected. This may be one reason long-term interest rates in the United States have fallen over the past two months.



    Almost ten years ago, the Federal Reserve began a quantitative easing (QE) program that more than quintupled the size of its balance sheet. At present, the Fed owns about $2.4 trillion worth of Treasury notes (11% of the total outstanding) and $1.7 trillion of mortgage-backed securities (19% of the total outstanding). The decision surrounding how much to return to private hands is therefore a potential market-mover, and there is considerable dispersion among analyst estimates of where the Fed will end up.

    The uncertainty is rooted in the fact that there is no precedent to rely on. In a sense, reducing the balance sheet is just another way of withdrawing reserves from the financial system to modulate credit, growth and inflation. But QE of this scale has no precedent, leaving no past data to help calibrate the optimal size and speed of the reduction.

    Further, the Fed has been struggling in its effort to manage the more traditional monetary lever of interest rates. To review, the Fed establishes a range for overnight interest rates, and then steers them from above and below. At the upper end is the interest rate on excess reserves (IOER) held by banks. Raising this rate offers banks a risk-free alternative to lending and (at the margin) should reduce the growth of credit.

    At the lower end of the target range is the Fed's reverse repurchase agreement (repo) program. In these transactions, the Fed uses its securities holdings as the basis for borrowing money from market participants for short periods (paying them interest for doing so). These arrangements remove reserves from the system, thereby raising overnight rates.

    After resting comfortably in the middle of the target range for most of the past ten years, overnight interest rates began trending towards the top last spring. And demand from investors for reverse repos has fallen from about $150 billion daily last October to nearly zero in recent weeks. These events suggest a paradox: reserves appear to be getting scarce, even though there remain trillions of dollars of them in the U.S. banking system.



    One reason for this might be that post-crisis reinforcement of bank capital and liquidity standards provided additional motivation for financial institutions to park money at the Fed. Central banks are low-risk counterparties, and balances held there count toward the liquidity coverage ratio (LCR) that financial institutions must maintain. Over the last year, increases in market rates have caused some bank depositors to drift back into investment products, heightening the need for banks to maintain liquid assets.

    Without a clear picture of banks' demand function for reserves, managing monetary policy becomes more complicated. In June, the Federal Open Market Committee set the IOER five basis points lower than the upper end of its target range for interest rates. This was an attempt to steer overnight interest rates down, and to a degree, it has succeeded. But banks are still paying a higher rate for short-term funding than the Fed would like to see.

    The shrinkage of the balance sheet has been slow so far; holdings are down just over $150 billion from their peak. The Fed has a set schedule for how the process will continue; declines of $40 billion to $50 billion per month are anticipated for the rest of 2018. Most think that there is a high bar for making alterations to the program in order to minimize uncertainty in markets.

    Few thought we might be approaching that bar so soon. But if signs of reserve scarcity persist, the Fed may have to reassess the neutral level of its balance sheet. If this level is raised, the central bank's securities holdings will not fall by as much as expected, which is one potential reason for the recent firming in Treasury note prices.

    The ideal level of the Fed's balance sheet was debated two years ago at the Fed's Jackson Hole economic symposium, and the debate is likely to be resumed next month at the 2018 edition. As much as the Federal Reserve would like to leave its balance sheet reduction program on autopilot, it may have to consider a manual override.